I want to know if I can access data stored in SQL database with data.table like dbplyr. Also, is there any package that can run regressions on out-of-memory data.
On PostgreSQL you can define functions with procedural languages, one of those is PL/R (also PL/pgSQL, PL/Tcl, PL/Perl, PL/Python, PL/v8, etc), so you can run regressions using R within the postgresql server itself.
Thanks for your reply. I am actually seeking a local solution instead of using a server.
Well, you can install postgresql server in your local system too
Can I do it in MonetDB? I use PostgreSQL every day, but it seems too slow for me.
I have never used monetDB so I don't know if it has a similar feature as procedural languages, also, I don't think that a well tuned postgresql installation could be considered as slow (compared with other sql servers) but tuning postgresql could be very difficult and system specific.
The biglm package generalizes the input for linear regression with the biglm() and bigglm() functions. The latter accepts SQLite or RODBC connections for the data argument. Both also accept a function which is called repeatedly to return chunks, so it can repeatedly grab a new chunk and update the model.
I've never used it, so I'm not sure of how efficient it is.
This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.
If you have a query related to it or one of the replies, start a new topic and refer back with a link.